Thread by @psybalazs

Despite the importance of #blinding in medical research, most trials don't assess blinding integrity, partially because there is no method to adjust trial results for blinding integrity... until now! New preprint with implications for #microdose and #psychedelic research 🧪🧵👇

First, we define activated expectancy bias (AEB), which is an uneven distribution of expectancy effects between treatment arms due to patients recognizing their treatment allocation. AEB can be viewed as residual expectancy bias not eliminated by the trial’s blinding procedure.

The main idea behind AEB is that if treatment allocation can be deduced by participants, then, treatment expectancy can bias the outcomes in the same manner as it biases non-blinded trials, for example as in open-label trials.

AEB is operative if two conditions are met: weak blinding (1) and when there is a positive trt expectancy (2) - #psychedelics are highly susceptible to both! We use computational modelling to show how AEB can inflate estimates of trt effects and create false positive findings.

To counteract AEB, we introduce the Correct Guess Rate Curve (CGRC), a novel analytical tool that can estimate what would be the outcome of a perfectly blinded trial based on data from an imperfectly blinded trial data

We apply CGRC to data from our self-blinding #microdose trial (tinyurl.com/52nnxp7d), showing that the observed acute placebo vs. microdose differences are likely to be false positives created by the combination of positive expectancy weak blinding (=AEB).

Based on these considerations, we #hypothesize that the claimed mood/cognitive benefits #psychedelic #microdosing can be understood as ‘active #placebo’, i.e., an intervention without robust direct benefits, but with recognizable effects.

A key insight is that we need to distinguish between 'trials with a placebo-control group', when a placebo group is formally present, and 'placebo-controlled trials', when the blinding is working as intended and participants are genuinely unable to decipher their trt allocation.

The implication is that placebo-controlled studies are more fallible than conventionally assumed, which has consequences for #evidencebasedmedicine and #publichealth policies.

My collaborators on this project were @ProfDavidNutt, @RCarhartHarris and @DavidErritzoe, thank you all for the great support 🙏! Preprint link: psyarxiv.com/cjfb6/

ps: would love to run this pipeline on other pharma trials, in particular SSRI trials. Maybe you can help? @JeremyHowick @eturnermd1 @Harvard_PiPS @LironRozenkrant @BenColagiuri @AmeliaJScott @Colloca_Luana @sfb_trr289 @kube_tobias @LSharpeUSYD @joannamoncrieff @MattBurkeMD